Paulo Blikstein is an Associate Professor at Teachers College, Columbia University, an Affiliate Associate Professor in the Department of Computer Science at Columbia University, and Director of the Transformative Learning Technologies Lab and of the Lemann Center for Brazilian Studies. Renato Russo is a doctoral student at Teachers College and a researcher at the Transformative Learning Technologies Lab.
Swiss cognitive scientist Jean Piaget demonstrated that there is nothing more resilient than a theory we create on our own. Narratives and stories are powerful, but they lack one crucial property, in comparison: they don’t make us feel as clever and intellectually capable. We propose that this pleasure and feeling of self-efficacy in theorizing – also proven by decades of neuroscience research – is closely related to current political communication and democracy-threatening events that took place in Brazil last January. Starting with the election of Lula in late October, thousands of Bolsonaro supporters spent as much as two months camping in front of military facilities, mobilized around the claim of rigged elections, culminating in the siege of the Brazilian capital on January 8th, 2023.
“Fake news” explains part of a campaign that elected Brazil’s Jair Bolsonaro four years ago and that kept part of his constituency mobilized. But it is only part of the story. Research by media studies and communication scholars, such as Francesca Tripodi, Alice Marwick, and Ethan Zuckerman, has shown how extremists resort to epistemological practices that in a way resemble those of scientific communities. Drawing on this scholarship, and combining it with classic constructivist theory, we propose a perhaps controversial explanation: many far-right voters are primarily theorists and avid learners who construct knowledge on their own.
Steve Bannon, whose strategies resonate among Bolsonaro and his aides – and not by chance the director of 19 documentaries – understood this clearly: instead of seducing or convincing voters, make them “learners,” “researchers,” help them to see “behind” appearances and “what media doesn’t show”, encourage them to do their own research and reach their conclusions. Bannon’s Brazilian disciples have structured a type of propaganda that does not impose the “truth” all at once, but strategically spreads small pieces of half-truths on social media that are then reconstructed – as solid, robust theories – in voters’ heads. Leaders come and go, and narratives last for weeks, but these self-learned theories can endure for decades.
We have been researching these far-right “teaching” strategies in Brazil since 2019, especially among Bolsonaro voters, and the data is staggering. Contrary to progressive imagination, adherence to the anti-democratic right is not always irrational. The most sophisticated contemporary online political propagandists don’t enunciate that “leftist governments always turn the country into new Venezuelas,” or that “sex education corrupts children.” They show examples, ask questions, describe cases. The voter notices patterns, sees parallels, draws conclusions, and feels proud to be a learner.
In our data, we often find phrases such as “I changed the way I thought due to YouTube videos”, “I understand the world so much better now”, and “Bolsonaro’s live videos explain what happens in politics” (emphasis added). The terms we highlighted are associated with knowledge and learning, not ignorance. Piaget reminds us that the construction of theories does not require those to be correct. It is enough that the pieces fit together approximately, generating the pleasure of seeing “further” than others. Once consolidated, even when based on delusional, racist, sexist, homophobic, and despicable principles, it becomes difficult to deconstruct those theories – especially if the learner is bombarded daily by “confirmation pills,” exploiting our proclivity to confirmation bias. This formula has worked like clockwork.
The deconstruction of these robust theories will be a task of years, or decades, in which we will have to put on the hat of educators and patiently offer theories that explain the world in a more coherent and ethical way. This will be the work of a generation, not just an election cycle.
To change theories (and votes) in the years to come, we need thus to understand the non-radical, “still-reachable” voter as an avid learner. To explore flaws in the spurious theories propagated online, we should not just present ready-made messages or insist on isolated fact-checking. If you have a strong theory, you can make almost any facts fit it (think pre-Copernican astronomy.) Instead, we need to ask questions, understand their theories, and deconstruct the flawed mechanisms in theory-building–not the facts themselves. For example, we could show that an isolated or extreme case cannot be turned into a rule, or alert for confirmation bias. And then find limitations in current theories, and co-create more subtle ones, seeking more data or alternative explanations, showing that there is intellectual satisfaction in changing how we think. The purpose is not to substitute one theory for another, but to show that the learning process in far-right networks is biased and guided, and that learning outside of them is richer, more interesting, and more rewarding.
In other words, voters need to unlearn how to learn in the closed ecosystem of the extreme right. This won’t always work. But we cannot give up. Each millimeter of this laborious change will be a mile towards the protection of democracy and civil society.